Dear all,

I have some questions about fully-differential amplifier simulation.
In the attachment,
fig1 is the schematic of the closed-loop amplifier setup.
fig2 is OTA DC Open Loop Gain(dB) vs. Vod(V).
fig3 is AC Loop Gain @Vod=0V.
fig4 is AC Loop Gain @Vod=2.1V.

Questions are:
1)How to build the testbench for these three simulations?
2)In spectre, typing "spectre -help stb", it will show "stability analysis of differential feedback circuits", cmdmprobe in analogLib library can be used to check the stability of the fully-differential feedback circuits just in fig1. So, how about the AC Loop Gain? Is the AC Loop gain based on Vod=0 or Vod=others? How to get the AC Loop gain@Vod=2.1V and @Vod=0V repectively?


Thanks.

CHMF,

  I would use the basic testbenches that you already have
and add:

1) The cmdmprobe so that you can perform stability analysis of
   the common mode and differential mode loops.

2) Add analysis dependent switches to define the dc operating
   point

3) For Vod=0V, I would directly connect the switch from input to
   output

4) For the Vod=2.1V, I would include a series voltage of +/-1.05V
   in the branch with the switches
   

   My assumption is that you mean loop gain, T(jw), when you say
AC Loop Gain. Stability analysis will return the loop gain.

                                                          Best Regards,

                                                             Sheldon